Still getting stutter

[DDogg]

Gee, I'll make sure to tell that to a lot of the people that had older players and this same problem before dvd-rb was a sparkle in your eye. I agree your techniques may have aggravated an existing problem, but it was, and is an existing problem. CCE will spike above 9000 when set at 9000, especially if the bias is set low. Older players with low xX drives have a problem dealing with the heavy near instant bitrate bitrate ramp caused by complex scenes like the ones mentioned in Revolutions. I'm happy to call that any barn animal you would like. Cow, horse or duck...it still stinks for the users of those older players.

[Video Dude]

It is generally a good idea to keep the max bitrate at 8000 if your target is recordable dvd (dvd-r or dvd+r). Some DVD players can't read recordables as "fast" as they can pressed DVDs, especially along the outer edges. It depends on both the quality of the media and DVD player

This is why the author of TMPGEnc limits the max bitrate on the DVD template to 8000, to prevent the stutter.

[DDogg]

I've attached a ScreenShot to this post of the information from the bitrate viewer in DVDLab. DVD-RB was not used to create this sample encode. It was a simple 2 pass Multipass VBR with 4500-0-9000 settings. Bias was set at 30.

Note the nearly instant ramp at the position of the cursor.

I'm always suspicious of any of the tools that show peak bitrate. None of them agree and I do hope we will get a reliable tool one day. What this does show, imo, is that peak bitrate in CCE is an area that needs more investigation. In the meantime, using a MAX of 8000 may be advisable.

/Add: the graph shown is using video only that has had pulldown applied. One or two audio tracks added to this would have really forced up the bitrate, even if only using the 9400 peak figure. I just don't know what to make of that 11k spike at the cursor position. Is it even real, or something misreported by DVDLab's bitrate viewer? The lower figure seems to agree with the standard bitrate viewer tool we are all familiar with.

[Video Dude]

EDIT: This is not a scientific test. It is not meant to reflect all media and players. Just my observations and experiences. I just wanted to point out that it may not be a problem with DVD Rebuilder but a problem with media or player causing the stutter.

Tested on:
2 brands of media
4 different brands of stand alone DVD players

Results:
MAX 9000: Stutter on all 4 players (throughout the disc)

MAX 8500: Stutter on 3 out of 4 players (at the end)

MAX 8000: No Stutter at all

[Sir Didymus]

Originally posted by Video Dude
I did some testing. My results show that recordable media can not handle bitrate over 8000 reliably...


Interesting...
Just for curiosity,
what is the brand and model of the players ?

[DDogg]

Video Dude, those are important and serious statements. I wonder if you could take the time to edit your post and further describe your testing methods? With DVD-RB? What bias if CCE? AVG/MinMax?, etc.,etc.. The important thing is having enough information to be able to duplicate your results. Scientific process and all that jazz :)

Btw, I think you are correct that 8000 is a good 'insurance' setting, although I have never thought about there being any difference between recordable media and pressed media. That's an interesting twist to the problem. I would like to learn more about it. All sources with more information gratefully received if you have any conveniently available.

Since these odd high bitrate spikes would only be caused in a VBR environment when the encoder encounters a highly complex scene (IMO), I don't think anybody would notice a whit of difference visually or quality wise using 8000 as a Max, although I freely admit to conjecture on the subject.

[Joergen]

Unreliable media on a low quality reader will surely drop the max read speed towards the end. If you expect that it reads at 1X, but add the read retrys and error correction it will be lower than this. Btw does a dvd spin at the same RPM at all times, so the reading gets harder as the data moves faster infront of the laser at the end?

[Video Dude]

Here you go DDogg, more details.

A few months ago I made a bunch of clips with my camcorder so I could test out the different mpeg encoders. I wanted the best settings for my vacation tapes. I intentionally made some extended complex scenes with lots of motion and zooming. I encoded them with TMPGEnc and CCE Basic. My Toshiba dvd player has an on screen bitrate viewer that overlays the bitrate while you watch the video. The high motion scenes required the most bitrate. I noticed when the high motion scenes has a bitrate higher then 8500 for more then 15-20 sec, the disc began to "thrash" in the player causing a stutter especially at the end of the disc. When the bitrate went down it played normally. I did some google research and found a possible cause might be the buffer in the dvd player. So I made the bitrate 8000 max and everything played perfectly. I know that commercial pressed discs go as high as 9800, so I assumed it must be due to the recordable. Then a while ago I read a posting on the official TMPGEnc boards and someone else noticed that high bitrates causes recordables to skip at the end, and said that is why the DVD template in TMPGEnc is set for 8000 (sorry I couldn't find the post).

This is no means a scientific test, it is just my opinion on why it stutters. I was reading this topic about the stutter and it reminded me of vacation tape testing. I just wanted to point out that it may not be a problem with DVD Rebuilder but a problem with media or player. I didn't mean to make a big deal about it to have you call it a serious statement. I'm sure there are many people not having problems with recordables.

[Joergen]

Originally posted by Video Dude
I noticed when the high motion scenes has a bitrate higher then 8500 for more then 15-20 sec, the disc began to "thrash" in the player causing a stutter especially at the end of the disc.

I've also done some DV to DVD and VHS to DVD discs with CCE and used 9000 for max and had no stutter problems in complex scenes. I've even done one disc with 9000kbit CBR without problems. This was on quality Ritek G03 media burned at 1X.

What audio did you use in conjunction with the video? I used 2ch 192kbit MP2.

[jdobbs]

Originally posted by Joergen
I've also done some DV to DVD and VHS to DVD discs with CCE and used 9000 for max and had no stutter problems in complex scenes. I've even done one disc with 9000kbit CBR without problems. This was on quality Ritek G03 media burned at 1X.

What audio did you use in conjunction with the video? I used 2ch 192kbit MP2. I'm also having a hard time with this. The standard associated with this is related to the spin rate of the disc (1x), pit size, and buffer requirements that lead to bitrate after including overhead, error detection, audio, etc as an end result... I don't think the disc can meet the minimum requirement to be labeled "DVD" if it couldn't handle this. The only way it should fail to meet it would be through errors and retries (as was mentioned by Joergen).

[Joergen]

Not to say Video Dude's findings arent true to his combination of encoder/burner/media/player.

Perhaps a "limit bitrate to 8000kbit" option in the expert settings would do well so people could try it or even use it to be "safe".

When you think about transcoding, DVDShrink always takes away so you'll never get anywhere near the max bitrate unless you compress by less than 90% of the original bitrate, which is rare.

I know its a hairy subject but what do the "groups" recommend for max bitrate for DVD-R? Those communities are experts at bitching and surely if they use 9000kbit or more for max it's ok for all :)

edit: a further question for Video Dude; Have you ever done a movie only or other 1:1 encode with dvdshrink of a movie that originally has 1 or 2 audio streams? (lots of streams means less video bitrate in the original) And do they work without stutter?

[Video Dude]

I think there was some confusion over what I posted. Let me clear it up so we can put this issue to rest.

I am not saying, or am I implying, or trying to argue recordable media can't handle a bitrate over 8000. I agree with jdobbs about the dvd standard point. My goal was to say perhaps the stutter "bug" was not the fault of DVD Rebuilder, but perhaps the media or dvd player. I only brought this issue up because reading this topic and seeing Ddogg's graph reminded me of my past experience. My stutter cleared up once I made 8000 max. I thought I would share my experience because I thought it might help somebody. Again, I did not mean that the universal max should be 8000, but those having the stutter problem might want to try 8000 to be safe. There are probably a much greater number of users not having problems then there are people experiencing the stutter. Perhaps the way I presented my test results, some misinterpreted what I was trying to say.

[DDogg]

Video Dude, very sorry if I inadvertently threw you to the wolves :) It was not my intention. This is the type of subject that pits the practical against the theoretical. I tend to opt for the practical. In this case, 'Revolutions' stutter on my old Apex when encoded with, or without dvd-rb, if the Max bitrate is 9000. It does not stutter if the Max bitrate is 8000. Guess which one I'll use :)

[Video Dude]

No problem, DDogg :D

[DDogg]

Originally posted by Joergen
I've even done one disc with 9000kbit CBR without problems. It darn well better work with CBR. Like jdobbs said, it is a dvd and and is supposed to handle rates like this. What I am speaking of is can it handle a total VBR bitrate spike of 12,000? 11,000? 10,000? Can all/any/some players go from near 0 bitrate to one of those rates is a few hundredths of a second? CBR should not exhibit any of these problems.

The 11k bitrate spike shown in DVDLab bitrate viewer graphic I posted troubles me (if accurate). The peak rate already showed a potential problem if enough audio tracks are added, but an 11k spike followed by many more of a similar rate would seem to nearly 'overload' the buffer and I would guess maybe the processing chips?

[DDogg]

Attached bitrate view of DV footage. Encoded OPV20 0-9000.

DV is really tough on an encoder.

@Mod - Moderate when convenient please.

[Joergen]

Originally posted by DDogg
What I am speaking of is can it handle a total VBR bitrate spike of 12,000? 11,000? 10,000?

If there really are peaks like that, it's surely a problem (unless commercial discs have lots of them too).

Beside the 1X reading speed for the DVD, there probably even exists a limit to how much data the mpeg2 chip or memory can pass through per second. Unless a specific DVD player's mpeg2 decoding chip is based on a DVB-S/T chip (that is supposed to handle up to 25mbit per second), but the cheapo brands most probably cut their hardware close to the limits to save in production costs.

Are those peaks limited to NTSC and caused by calculation errors?

[jdobbs]

The maximum peak a player can stand depends a lot on buffer size.

[Obi Wan Celeri]

Whoo it's been a long time :)

Has anyone ever noticed that the actual .M2V file was stuttering?
I was trying to make a personnal backup of the fantastic "Allegro Non Troppo" (a must if you like animation and classical music) and after burning it and watching it on my home DVD player I noticed stuttering in 3 chapters (of an extra on the DVD).

So I bought a brand-name DVD-R and burned ... same problem.

I finally checked the source and discovered that the actual .M2V file was jumpy ... So far it's only done this to me with this particular DVD.

Searching further I found that the .AVS file also produced a stuttering video ... so maybe DVD-RB has a problem "cutting" at chapter ends in some specific cases?

It's also possible that the original DVD had the chapters end at a weird place (certainly seems like this here).

I might try to manually fix the .AVS files and try forcing DVD-RB to reassemble the whole thing afterwards ...

For those who would like to know here are my specs ...
Allegro Non Troppo NTSC
DVD Rebuilder 0.49
DVD Decrypter 3.2.2.0
AVISynth 2.54, Jan 26 2004 [09:28:02]
MPEG2Dec3dg.dll 1.0.1.0
Decomb 5.10
CCE SP 2.66.1.7
Pioneer A06

Thanks!
Obi Wan Celeri

[Obi Wan Celeri]

Fixed the stuttering by changing the following line in one of the .AVS files ...

Before
trim(21502,31537)

After
trim(21501,31537)

I moved the frame back by 1 ...

Now this is all fine and dandy but is there a way to fix this automagically? Now that's the 1,000,000$ question ...

Obi Wan Celeri
D2V file available on demand (more if needed)

[Sir Didymus]

Originally posted by Obi Wan Celeri
Fixed the stuttering by changing the following line in one of the .AVS files ...

Before
trim(21502,31537)

After
trim(21501,31537)

I moved the frame back by 1 ...

Now this is all fine and dandy but is there a way to fix this automagically? Now that's the 1,000,000$ question ...

Obi Wan Celeri
D2V file available on demand (more if needed)

:confused: :confused: :confused:

What do you mean with this ?
Did you verify carefully that the original cell contain one frame more than the rebuilt cell, before doing your test ?
Did you discover that Mpeg2Dec3dg.dll is actually decoding a wrong number of frames ?
Did you just perform some "let's try and see..." test with your avs files ?

Please consider that some badly produced original DVD may "stutter" or even not play at all due to errors or scratches...

[Obi Wan Celeri]

Originally posted by Sir Didymus
:confused: :confused: :confused:

What do you mean with this ?
Did you verify carefully that the original cell contain one frame more than the rebuilt cell, before doing your test ?
Did you discover that Mpeg2Dec3dg.dll is actually decoding a wrong number of frames ?
Did you just perform some "let's try and see..." test with your avs files ?

Please consider that some badly produced original DVD may "stutter" or even not play at all due to errors or scratches...

Ok sorry if I wasn't clear. Hopefully this reply will straighten things out a bit :)

First of all the original DVD works properly on my DVD player. The disc is not scratched at all. Nor does it skip when played. Nor is it a bitrate issue; theoretically such bitrate bottlenecks don't exist on a PC when playing off a hard drive.

The extracted VOBs are equally OK; they play properly too. No skipping (used Media Player Classic and opened the main IFO).

So logically, DVD-RB has something to do with the problem. Let's proceed step by step.

1. A few chapters of the final (converted) movie stutter (even when played directly from the hard disk). Since we know the source is good, then it must be because of DVD-RB / CCE / Decomb.
2. The corresponding .M2V stutters. The problem didn't happen rebuilding the DVD (last phase).
3. Searching deeper, the corresponding .AVS file, when played in Virtual Dub, stutters there too. So CCE has nothing to do with this problem (Encoding phase).
4. Activating or deactivating Decomb doesn't seem to make any diference. So bar that too.
5. Only by manually changing the corresponding .AVS file (as per my previous example) can we get the stuttering to dissapear.

We are therefore left with a few possible causes for this problem:
a. DVD-RB has a bug and cuts chapters/cells at the wrong place
b. The original DVD cuts chapters/cells at the wrong place
c. Something is wrong with me and I goofed up somewhere.

We can safely rule out b. since the DVD and the extracted files run perfectly.
While c. is possible, I made sure I checked everything before posting here.
Which leaves a. as the only possible source.
And as I said this might end up being quite a nasty little problem; how is it possible to identify frames that will cause stuttering if played first? Also being forced to manually edit .AVS files is not something really user friendly ;)

So now hopefully this is a bit clearer :)

Obi Wan Celeri

[Sir Didymus]

Ok.

Please consider the following is just related to the video contents...

Maybe there are other simpler and more efficient ways, but the process I would use to verify if the cutting points produced by DVD-RB are right, would be to open the the original VOB with vobedit and demux it by cell id. You get a lot of smaller VOB, each one corresponding to the cellid/vobid of the title. You open again with vobedit the specific vob you want to analyze and demux its video stream, producing the related m2v file.

Now, for the specific cell you want to look at, you have two m2v: one is extracted directly from the original and the other is the one produced by DVD-RB. Both should have the same frame number and scene changes position. Using dvd2avi and virtualdub you should be able to get a deeper look at the m2vs for obtaining a definitive answer to item a. in your post... :)

Good look and cheers
SD

[Toranaga]

Ok, here I thought DVD-RB was ready to take on my Bond collection. But after doing 6 films, i realize that there is still a problem with the audio drop-outs :confused:

I thought this had been solved? I am using 0.49.

[jdobbs]

Originally posted by Obi Wan Celeri
Ok sorry if I wasn't clear. Hopefully this reply will straighten things out a bit :)

First of all the original DVD works properly on my DVD player. The disc is not scratched at all. Nor does it skip when played. Nor is it a bitrate issue; theoretically such bitrate bottlenecks don't exist on a PC when playing off a hard drive.

The extracted VOBs are equally OK; they play properly too. No skipping (used Media Player Classic and opened the main IFO).

So logically, DVD-RB has something to do with the problem. Let's proceed step by step.

1. A few chapters of the final (converted) movie stutter (even when played directly from the hard disk). Since we know the source is good, then it must be because of DVD-RB / CCE / Decomb.
2. The corresponding .M2V stutters. The problem didn't happen rebuilding the DVD (last phase).
3. Searching deeper, the corresponding .AVS file, when played in Virtual Dub, stutters there too. So CCE has nothing to do with this problem (Encoding phase).
4. Activating or deactivating Decomb doesn't seem to make any diference. So bar that too.
5. Only by manually changing the corresponding .AVS file (as per my previous example) can we get the stuttering to dissapear.

We are therefore left with a few possible causes for this problem:
a. DVD-RB has a bug and cuts chapters/cells at the wrong place
b. The original DVD cuts chapters/cells at the wrong place
c. Something is wrong with me and I goofed up somewhere.

We can safely rule out b. since the DVD and the extracted files run perfectly.
While c. is possible, I made sure I checked everything before posting here.
Which leaves a. as the only possible source.
And as I said this might end up being quite a nasty little problem; how is it possible to identify frames that will cause stuttering if played first? Also being forced to manually edit .AVS files is not something really user friendly ;)

So now hopefully this is a bit clearer :)

Obi Wan Celeri You've ignored the impact of AVISYNTH and MPEG2DEC3. Both are still players at the point you are discussing. Tell me this... what is the frame numbers of the AVS immediately preceding this one? If it ends at 21501 you've missed the mark. I'm betting it does. If not, we are onto something.

[Obi Wan Celeri]

[Toranaga]

Actually, in my case it's more of a "video stutter". The audio is, for all intents and purposes perfect. In fact DVD-RB doesn't touch the audio at all, it just transfers it.

BTW, If anyone would like me to I could show pictures of what happens (I'd post 5-10 frames between two chapters) when the stuttering starts.

Anyways, to be completely honest, DVD-RB has been fantastic in 8 of the 10 titles I've thrown at it. My guess is you can easily do all your Bond collection - just watch them all to be sure ;)

... my name is celeri, creme of celeri :cool: ...

Oh, you can try different software but so far DVD-RB gives the best results.

[Sir Didymus]
Thanks for the info. I will look into it when I get a chance :)
And I will (of course) post my findings on this thread!

[Obi Wan Celeri]

[JDobbs]
I'm definetly NOT an expert but if I understand you correctly you are telling me the problem lies within MPEG2DEC3 or AVISynth; maybe they have a problem with this particular stream. Maybe.

Ok, as far as the frame sequence is concerned, here are the contents of two subsequent .AVS files. It all looks perfect on paper but on practice I get jerkiness.

#------------------
# AVS File Created by DVD Rebuilder
# VOBID:01, CELLID:03
#------------------
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\Decomb510.dll")
mpeg2source("C:\ALLE\D2VAVS\V03.D2V")
trim(19032,21501)
FieldDeinterlace(blend=false)
ConvertToYUY2(interlaced=true)

<--- This is where the jerkiness starts --->
<--- The first 100 frames are in the clip below --->

#------------------
# AVS File Created by DVD Rebuilder
# VOBID:01, CELLID:04
#------------------
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\Decomb510.dll")
mpeg2source("C:\ALLE\D2VAVS\V03.D2V")
trim(21502,31537)
FieldDeinterlace(blend=false)
ConvertToYUY2(interlaced=true)

And I also made a small video that corresponds to frame 21502-21601 in DivX 5, no audio (audio isn't an issue here). Notice the blocks and the jerkiness.
http://www.celeri.org/Video_stuttering_in_allegro_non_troppo.avi

Hopefully this is not caused by a stupid error on my part ;)
P.S.: I might try decoding with IDCT=7 ... who knows?
[EDIT]: Tested ... and it didn't work :(

[jdobbs]

I wasn't implying it was AVISYNTH or MPEG2DEC3 -- I was pointing out that you had not yet ruled out all possiblities as you had said. Also you had saidDVD-RB has a bug and cuts chapters/cells at the wrong place The point I am making is that this is not likely because of the thousands of chapters it has calculated split successfully. It is also unlikely because of the way I know the code works -- the fact that the two files have different VOBID/CELLID combinations means that there was a NAVPACK between the two frames that changed the IDs. The AVS files you just posted seems to substantiate what I'd said. If you change the start frame in the second AVS to 21501 (as you said you did) -- you are repeating that frame in both cells. I can assure you that isn't correct.

The example you've given doesn't actually "stutter." It just shows some blocking artifacts. I couldn't tell from the example, but it could be an authoring error on the original. My guess is that the second cell in your example has frames that are not temporally accurate or complete. If the point of the blockiness is a cut point between two cells, it certainly is a strange place to have one.

I'm not really sure what else to tell you on this one. Sorry.

[AzraelsCurse]

After reading Obi Wan Celeri's posts, I checked the m2v's of the movie's that I get stutter in. I notice that they do not stutter when played on the computer, but I did notice that the stutter(seen only in stand-alone's) is at the beginning of an m2v. It can be anywhere in a chapter, but I thought it was intresting that it was at the beginning of a new m2v(every time).
I'm going to test authoring a dvd using several of the seemingly trouble m2v's, and if they don't stutter, maybe some stand-alone's don't like something with how DVD-RB is rebuilding the dvd from the m2v's?
Does anyone else who is having stand-alone only stutter notice that it is at the beginning of m2v's?

Update:
I grabbed some of the stuttering(stand-stand alone only) m2v's from 2 different dvd-rb projects, merged them(each dvd's m2v's merged separately) added pulldown, created a dvd with tmpgenc dvd author(without audio, or chapter breaks, ect.) and it played without stutter on my stand-alone dvd players. So, that seems to say that the encoding phase works correctly, and since the encoding phase relys on the prepare phase, then the prepare phase is at least mostly working correctly, so I'm leaning toward a rebuild phase problem. I'm certainly no expert though, and I certainly don't fully comprehend the full level of Magic that jdobbs is working here, so maybe some of the experts can make something out of this information. But who knows, maybe I'm the only one still getting usually minor stutter when playing back DVD-RB back-ups in my stand-alone dvd players.
I have tried the min_bitrate and max_bitrate settings as recommended in this thread and it had no effect on the stutter that I am getting, however I only tested one dvd so far with those settings.

[DVD Maniac]

I tried Master and Commander with Rebuilder, keeping the DTS sound track. When played on both my set-tops it stutters immediately at the start of the film and continues every 10 seconds or so. I ran the offending vob file through bit rate viewer and it was only peaking at 6500 so I just don't understand why its doing this.

I also tried Intolerable creulty (again with DTS) - same result but this time its peaking at around 8500.

Why this inconsistency? From my experience, Rebuilder just does not seem to cope with DTS at all.

What bitrate settings should I be trying in the setup file to get DTS encodes working?

[Obi Wan Celeri]

Originally posted by jdobbs
The example you've given doesn't actually "stutter." It just shows some blocking artifacts. I couldn't tell from the example, but it could be an authoring error on the original. My guess is that the second cell in your example has frames that are not temporally accurate or complete. If the point of the blockiness is a cut point between two cells, it certainly is a strange place to have one.

I'm not really sure what else to tell you on this one. Sorry.

First of all, thanks for looking into this :)
No matter if it can't be resolved. Let me know if you need me to run some extra tests.

Well this blockfest goes on for three chapters and then subsides and everything is back to normal. The more I look at it, the more I think the cut is on a frame that should'nt be cut. As for the fact that it doesn't look like stuttering, the next chapter is a bit more evident but still it does give a pretty good idea.

The material is also half video and half animation, on top of which it's PAL stuff converted to NTSC. Certainly ripe for errors. Although I must say the conversion is nice and clean and the jump between chapters is allright (when watching the original).

Other than that I'm really sorry it takes me forever to reply ... keep ... on falling ... asleep on my ... keyboard (play Star Trek action theme)

Obi Wan Celeri

[Joergen]

Celeri's video clip looks most like DV-errors in the master tape. Boy were those people blind (and dirty to make their tapes so terrible) :p

In any case it is definetly not MPEG2 errors since MPEG2 erros show themselves as various RGB error blocks (usually green) and as smaller blocks. DV-errors are clean and made of large blocks of the previous unerrorous blocks.

[Noah]

Version 0.50

Below are a summary of the changes/updates associated with this version:

- Corrected an error in which the maximum bitrate could peak higher than that set in the max_bitrate setting when working with interlaced sources. This is highly likely to have been the source of infrequent frame stutters at high demand points in encoding. The problem is the result of DVD-RB's inherent method of feeding frames to CCE and QuEnc. Please note that if you browse the .ECL files for interlaced sources, they will now show a vbr_brate_max= value that os 80% of the default setting -- do not change it. It is right, and the resulting stream (after processing) will still match the max value. Excellent observation by Trahald.
I [and it seems others] have had this problem with progressive sources. Has this been addressed for progressive frames as well or should we continue to manually set vbr_brate_max=8000? I guess I'll test the new version on what I just encoded (of course, forgetting to lower vbr_brate_max ;)) and report back.

EDIT: Looks good. Version 0.50 is setting vbr_brate_max=7200 on a progressive source, which results in a MAX/AVERAGE bitrate of 8792/7648 from a 9209/7553 source at 0% reduction. vbr_brate_max=8000 on the same clip produces 9471/7654. I'll think I'll be leaving it at 7200 unless I'm only doing very limited compression, in which case I'd probably be using a transcoder anyway.

Nice work, jdobbs! Might want to update the version history to show that this is fixed for progressive sources, as it seems to be.

[TheBigDave]

I tried the latest DVDRB (v0.50), and I'm still getting the brief stutter at abnormally high bitrates. Putting in the "max=8000" has worked well. But the new release hasn't fixed my problem.

Here are some Bitrate Viewer pics of a stuttering scene:

Original (no stutter):

http://images.andale.com/f2/115/106/1203959/1087213796400_Original.jpg

DVDRB v0.49 with max=8000 (no stutter):

http://images.andale.com/f2/115/106/1203959/1083927100469_RE_NoStutter.jpg

DVDRB v0.50 without max=8000 (stutter):

http://images.andale.com/f2/115/106/1203959/1087227959922_RE_Stutter.jpg

As you can see, with DVDRB v0.50 (without max=8000) I'm still getting a 10,000+ bitrate spike. This particular stutter is not at the beginning or end of a cell. It's right in the middle of one.

Anyways, I'm happy with using DVDRB and the "max=8000". That works fine for me, and I haven't had any problems with it.

If I can provide any additional information for you, please let me know.

[DVD Maniac]

Thanks for this info BigDave

Further to my earlier post on this subject, I am getting frequent stuttter problems that do not appear to be consistently linked to high bit rates, specifically when retaining DTS sound tracks. I have tried the max bit rate = 8000 setting with v0.49 last night but still with stutter problems with the DTS track retained. Odd thing is that I am getting stutter with DTS tracks with bit rates peaking well below the so called stutter threshold (only around 6500 peak).

I am going to try v0.50 with / without the max rate setting to see what happens.

Anyone any ideas why DTS seems to be such a problem for Rebuilder?

[the-warriners]

Forgive my ignorance but where do I set the max_bitrate in DVD-RB

I have been through all the menus and I cant see the setting anywhere

[DVD Maniac]

You need to put the settings in the Rebuilder.ini file using notepad or equivilant editor.

Try these settings as recommended by ddog -

[Options]
min_bitrate=300
max_bitrate=8000

[CCEOptions]
VBR_bias=33

[TheBigDave]

Originally posted by the-warriners
Forgive my ignorance but where do I set the max_bitrate in DVD-RB
I have been through all the menus and I cant see the setting anywhere

Where your Rebuilder.exe file is located you'll also see a rebuilder.ini file. Open it up. You'll find sections for [Options], [Setup], [Paths], [CCEOptions] and [Audio].

Under [Options] add:

min_bitrate=300
max_bitrate=8000

And under [CCEOptions] change the VBR_bias to read:

VBR_bias=33

[Toranaga]

There is no [Audio] section in the ini file.

[TheBigDave]

Originally posted by Toranaga
There is no [Audio] section in the ini file.

I have one in mine. It says:

[Audio]
VTS01=11000000

[Noah]

DVD Maniac,

Are you sure your problems are with DTS? Have you tried doing DD5.1 equivalents?

I've had some similar problems. I did Die Another Day (among others, which I haven't completely checked) with DTS before the stuttering/bitrate peak issue became known, and I ended up with stutters at nearly every chapter break and a handful of 10,000+ bitrate peaks. Maybe 40 across the whole movie, so not on the order of every 10 seconds, but I'd sure like to know I can keep the best audio track without expecting this.

Meanwhile, Daredevil with DTS made it through without the chapter break problems and only 1 bitrate peak that I've found.

Both of these were reauthored with Shrink to movie-only before re-encoding, so the authoring may not be perfect, either.

Obviously, we need to be checking DD5.1 versions, preferably at the same % reduction rate, or maybe just keep both tracks. Is it possible to actually select audio tracks after encoding in 3-click mode? I seem to remember that at least being selectable...whether it would actually change anything I'm not sure, but I can't see why not. If so, audio tracks could be rebuilt into the same encoded video. That would be the best way to isolate DTS as problematic.

[DVD Maniac]

Noah,
I have tried the equivilants with dd5.1 and the stutter problem goes away. Bitrate analysis of the resulting failed outputs (with DTS) from Rebuilder are not consistently peaking over 8000 but yet still the stutter problems - thats why I think its specific to DTS.

I am going to run some tests with v,050 tonight to see if this version solves the problems.

[Noah]

Originally posted by DVD Maniac
Noah,
I have tried the equivilants with dd5.1 and the stutter problem goes away.Bitrate analysis of the resulting failed outputs (with DTS) from Rebuilder are not consistently peaking over 8000 but yet still the stutter problems - thats why I think its specific to DTS.
Hmmm....that's convincing enough. Can you determine if the stutters correspond to any characteristic of bitrate? Maybe a very low bitrate or a wild swing in bitrate, in which case raising the VBR_Bias might be worth a try. I just can't think of what to check beyond that, since I don't think DVD-RB does anything with the audio stream apart from demuxing and remuxing. I'm really not familiar enough with how that works to troubleshoot those processes.

I am going to run some tests with v,050 tonight to see if this version solves the problems.
Don't get your hopes up...AFAIK, the only fix in 0.50 is to set vbr_brate_max=7200 (or 0.8 * max_bitrate if you override the 9000 default).

BTW, what DVD player(s) have you tested this on?

[DVD Maniac]

Noah,

DVD Players tested are Philips DVR70 and Panasonic 717 (the later normally VERY forgiving of DVD compliance requirements, the former is a real pig for EXACT compliance)

Lets see what v0.50 comes up with with a DTS tonight run ....

[Toranaga]

Just ran a Bond movie with 0.50 and the custom settings and it seems it helped with the audio stutter.

[DVD Maniac]

I ran a first test with a DTS title that was stuttering with v0.49 and the problem seems to have been fixed. I want to run some more tests with some other DTS titles as I am not convinced that bit rate spikes were the only problem.

So far so good though!

[Toranaga]

Originally posted by TheBigDave
I have one in mine. It says:

[Audio]
VTS01=11000000

I don't have this setting. Do I need it?

[AzraelsCurse]

Tested 2 previously stuttering movies with .50...
It has completely removed the stuttering in Runaway Jury, and slightly reduced the minor stuttering in Armageddon.
However the 2 spots in Armageddon that stutter in my Sony, continue to barely, yet noticably stutter. Overlookable even(it's only in the sony). It's a shame it's not silky smooth(yet?) like a IC8, but the quality is just unbeatable.
When I previously tried forcing max bitrate in Runaway Jury, it either did nothing, or 8000 was still too high. .50 backs it up perfectly though.

[jdobbs]

Originally posted by TheBigDave
I have one in mine. It says:

[Audio]
VTS01=11000000 The [Audio] section is used for saving selected audio exclusions in .RBD (project) files. When a project file is loaded, it's settings are put in the .INI file. So, if you never use open/save project options -- it won't be there. It also isn't read from the INI unless a "Open Project" has taken place.

The 1's and 0's each represent one of the audio streams in the "Audio Streams to Keep" list box. 1=Keep, 0=Don't keep.

[DVD Maniac]

Well I (tried) watching two results from Rebuilder v0.50 last night and I am afraid to say the stuttering problems are still there. Here is the bitrate data -

1. Master and Commander (Progressive) with DTS retained

Source max bit rate 8300
Rebuilder v0.50 6588

2. Kill Bill Vol1 (Interlaced) with DTS retained

Source max bit rate 8266
Rebuilder v0.50 7730

In the first case I was getting stutter about 10 times through the film, the second one was as bad as with v0.49 with stutter just about constant through playback. Switching to my more forgiving set-top pretty much eliminated stutter from the first title and reduced it slightly on the second.

Looking at the data above there seems to be a correlation between increased stutter and higher max bit rates, BUT -


Why is there still stutter at all, even though the resulting encodes are below the 8000 threshold?

I still think this is a DTS related problem with Rebuilder as DD5 encodes work just fine with Rebuilder v0.50 and the above examples